A dataset in relation to a three-dimensional object may be generated by capturing images of the object with an image capturing device from a plurality of angles. Such a dataset may include features extracted from images of the object captured from different angles and may be useful in a number of applications. For example, the dataset may be used to detect the presence of the three-dimensional object in an Augmented Reality application. The dataset as described above may be used by the Augmented Reality application or an associated application to determine, e.g., the pose of the three-dimensional object in the camera view.
As described above, to generate the dataset, images of the object are taken from different angles. Guidance and feedback information in relation to whether images of the object has been captured to a satisfactory degree at a particular angle may be helpful to a user that is capturing the images of the object from different angles. Images of the object has been captured to a satisfactory degree at a particular angle when the resulting dataset can be used for a particular purpose for the particular angle with a predetermined success rate or better. One such purpose may be, for example, to detect the three-dimensional object and its pose in a camera view at a particular angle using the dataset. Further guidance in relation to the quality of the captured information for a particular angle and/or a particular distance may also be helpful. Known methods may require the user capturing the images of the object to have an intimate knowledge regarding feature extraction and related processing. A method that provides more intuitive guidance that is less demanding on the user may be useful.